Content uploaded by Lorraine Hope
Author content
All content in this area was uploaded by Lorraine Hope on Oct 20, 2015
Content may be subject to copyright.
Law and Human Behavior
Memory and the Operational Witness: Police Officer
Recall of Firearms Encounters as a Function of Active
Response Role
Lorraine Hope, David Blocksidge, Fiona Gabbert, James D. Sauer, William Lewinski, Arta
Mirashi, and Emel Atuk
Online First Publication, October 5, 2015. http://dx.doi.org/10.1037/lhb0000159
CITATION
Hope, L., Blocksidge, D., Gabbert, F., Sauer, J. D., Lewinski, W., Mirashi, A., & Atuk, E. (2015,
October 5). Memory and the Operational Witness: Police Officer Recall of Firearms
Encounters as a Function of Active Response Role. Law and Human Behavior. Advance
online publication. http://dx.doi.org/10.1037/lhb0000159
Memory and the Operational Witness: Police Officer Recall of Firearms
Encounters as a Function of Active Response Role
Lorraine Hope
University of Portsmouth David Blocksidge
Metropolitan Police, London, United Kingdom
Fiona Gabbert
Goldsmiths, University of London James D. Sauer
University of Tasmania
William Lewinski
Force Science, Mankato, Minnesota Arta Mirashi
University of Portsmouth
Emel Atuk
University of Portsmouth
Investigations after critical events often depend on accurate and detailed recall accounts from operational
witnesses (e.g., law enforcement officers, military personnel, and emergency responders). However, the
challenging, and often stressful, nature of such events, together with the cognitive demands imposed on
operational witnesses as a function of their active role, may impair subsequent recall. We compared the
recall performance of operational active witnesses with that of nonoperational observer witnesses for a
challenging simulated scenario involving an armed perpetrator. Seventy-six police officers participated
in pairs. In each pair, 1 officer (active witness) was armed and instructed to respond to the scenario as
they would in an operational setting, while the other (observer witness) was instructed to simply observe
the scenario. All officers then completed free reports and responded to closed questions. Active witnesses
showed a pattern of heart rate activity consistent with an increased stress response during the event, and
subsequently reported significantly fewer correct details about the critical phase of the scenario. The level
of stress experienced during the scenario mediated the effect of officer role on memory performance.
Across the sample, almost one-fifth of officers reported that the perpetrator had pointed a weapon at them
although the weapon had remained in the waistband of the perpetrator’s trousers throughout the critical
phase of the encounter. These findings highlight the need for investigator awareness of both the impact
of operational involvement and stress-related effects on memory for ostensibly salient details, and reflect
the importance of careful and ethical information elicitation techniques.
Keywords: eyewitness memory, stress, arousal, law enforcement, interviewing
Law enforcement officers, military personnel, and others in civil
or emergency response occupations are frequently involved in
dynamic, challenging incidents. Depending on their particular op-
erational mandate, these “operational witnesses” may need to act
to preserve life, protect citizens, neutralize threats, initiate recov-
ery, or engage in some combination of related activities to resolve
an incident. Accurate and detailed accounts of the incident, and
information about the operational witness’s own activities and that
of colleagues, may be important for subsequent investigations and
the eventual delivery of justice (Alpert, 2009). Nowhere is this
more critical than in the case of shootings by armed police officers
who are authorized by the State to discharge a weapon in the
course of their duty to protect and avert imminent threats to life
(ACPO [Association of Chief Police Officers], 2011; Armed Po-
licing Authorised Professional Practice, 2013). The current re-
search examined the effects of active involvement on eyewitness
recall memory. We compared the recall performance of operation-
ally active witnesses responding to a (simulated) threatening inci-
dent with that of nonoperational witnesses, or bystanders, to the
same event; hereafter referred to as active and observer witnesses,
respectively.
Recent high profile cases, such as the shooting of Mark Duggan
in the United Kingdom, serve to highlight tensions in the investi-
gation of armed incidents. Duggan was shot by an armed officer in
London. A Public Inquest, conducted between September 2013
Lorraine Hope, Department of Psychology, University of Portsmouth;
David Blocksidge, Metropolitan Police, London, United Kingdom; Fiona
Gabbert, Department of Psychology, Goldsmiths, University of London;
James D. Sauer, School of Psychology, University of Tasmania; William
Lewinski, Force Science, Mankato, Minnesota; Arta Mirashi and Emel
Atuk, Department of Psychology, University of Portsmouth.
Correspondence concerning this article should be addressed to Lorraine
Hope, Department of Psychology, University of Portsmouth, Portsmouth
PO1 2DY, United Kingdom. E-mail: lorraine.hope@port.ac.uk
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Law and Human Behavior © 2015 American Psychological Association
2015, Vol. 39, No. 5, 000 0147-7307/15/$12.00 http://dx.doi.org/10.1037/lhb0000159
1
and January 2014 concluded that the shooting had constituted a
lawful killing (see http://dugganinquest.independent.gov.uk/latest-
news.htm). However, in the course of both the investigation and
inquest, questions were raised with respect to officers’ accounts of
the incident, and perceived inconsistencies in the statement pro-
vided by the officer who discharged his weapon were widely
reported in the media. Perhaps unsurprisingly, skepticism over the
accounts provided by officers in the aftermath of fatal shootings is
well documented both in the U.K. and internationally, and gener-
ally focuses on the potential for police collusion or corruption
(Braidwood, 2010; Heaton-Armstrong & Wolchover, 1993, 2009;
see also Hope, Gabbert, & Fraser, 2013). Although the deliberate
fabrication of evidence lies beyond the scope of the current article,
there is a less controversial explanation for at least some of the
apparent inconsistencies and inaccuracies in the “honestly held”
accounts provided by operational responders in the aftermath of
stressful incidents. The common lay belief that memory operates
like a video-recorder—often subscribed to by investigators, legal
experts, and potential jury members—is woefully inaccurate (Si-
mons & Chabris, 2011; see also Benton, Ross, Bradshaw, Thomas,
& Bradshaw, 2006). Memory is a dynamic and reconstructive
process, susceptible to error and distortion (Schacter, 1999). It is
well documented, in over 40 years of research, that eyewitness
memory is fallible—even under optimal encoding conditions
(Conway, 2012).
The recall of operationally active witnesses may suffer addi-
tional performance decrements because of their role in challenging
response contexts; for instance, as a result of increased cognitive
demands associated with response generation or decision-making.
Understanding how memory performs under such conditions is
important for a number of reasons. First, there is a dearth of
research examining memory performance under realistic opera-
tional conditions (simulated or otherwise). Second, it is important
that procedures and policies relating to the treatment of operational
witnesses by independent investigators or other agencies are (a)
well-informed with respect to the malleability of memory and (b)
appreciate the necessity for careful and ethical memory elicitation
practices. Third, it is critical that practitioners in the legal system,
whether lawyers, prosecutors, or judges, are cognizant of the
potential for naturally occurring memory errors, gaps, and incon-
sistencies among all witnesses but particularly operational wit-
nesses, such as armed officers, who are often expected to provide
detailed and accurate accounts of challenging incidents they were
immersed in.
In light of these concerns, we examined the memory perfor-
mance of witnesses with operational duties (i.e., police officers
actively responding to a simulated incident) with that of witnesses
who had not been operationally deployed (i.e., observers). We also
explored the effect of schema-driven expectation in operational
contexts on memory error.
Stress, Arousal, and Memory Performance
Armed officers regularly find themselves in unpredictable, dan-
gerous environments and may experience varying degrees of emo-
tional arousal and/or stress response (Meyerhoff et al., 2004).
Research on the effect of arousal on cognitive processes in applied
training settings reveals the cognitive and memory difficulties
experienced in high-stress environments. For example, Morgan
and colleagues (2004) tested the memories of soldier participants
who had been exposed to high levels of interrogation stress,
including physical confrontation, in an intensive survival school
training exercise. Memory performance, in terms of the recogni-
tion of a target individual who had physically confronted and
threatened them for over 30 min, was impaired after high stress
(vs. low stress) interrogations. More recently in a study involving
861 soldiers in a similar survival training simulation, Morgan,
Southwick, Steffian, Hazlett, and Loftus (2013) observed that
memories for stressful events, like memories for more mundane
events, are susceptible to misleading postevent information. Chal-
lenging field environments are also associated with significant
impairment in selective and sustained attention (Leach & Ansell,
2008) and reduced working memory capacity (Leach & Griffith,
2008). Focusing on the performance of police officers in simulated
operational settings, Hope et al. (2012) examined the effects of
physiological stress, as a function of exertion, on recall and rec-
ognition and found that police officers who had been exerted
reported significantly fewer correct details about an encounter, and
were significantly less likely than nonexerted officers to identify
an encountered target individual.
The precise mechanisms underpinning memory impairment as a
result of stress and arousal are difficult to directly delineate in the
applied context. A large body of literature confirms that emotion-
ally arousing events are remembered better than neutral events
(e.g., Payne et al., 2006) with neurobiological research, in partic-
ular, suggesting that stress hormones can enhance memory con-
solidation (McGaugh, 2013; Roozendaal, 2000). Researchers have
also speculated that attentional narrowing under arousal underpins
this recall advantage such that memory for central or important
stimuli is enhanced (on the grounds that such items “capture
attention”) while memory for peripheral items is impaired (Safer,
Christianson, Autry, & Österlund, 1998; see also Easterbrook,
1959). However, there are a number of problems applying this
rather simplistic account to the complex interaction between stress
and arousal on memory in applied contexts. First, the detrimental
effects of high levels of stress experienced in naturalistic settings
have been well-documented (e.g., Morgan et al., 2004, 2013). In
fact, in their meta-analysis, Deffenbacher et al. (2004) identify
what they describe as a “catastrophic” decline in memory perfor-
mance at higher stress levels. As such, the effect of arousal on
memory performance reflects an inverted U-shaped curve with
memory for events best when stress levels are moderate (Morley &
Farr, 2012). Thus, while arousal may activate the amygdala (Adol-
phs, Tranel, & Buchanan, 2005; Phelps, 2006), higher levels of
stress work to disrupt hippocampus function, impairing memory
for sensory detail and visuospatial working memory (Shackman et
al., 2006; for an extended version of this argument see Davis &
Loftus, 2009). Furthermore, pharmacological research observes
that stress hormones in the form of glococoricoids and cat-
echolamines (adrenaline/nor-adrenaline), naturally released during
stress (De Kloet et al., 1998), have variable effects on memory,
depending on a number of modulatory factors (Lupien & Lepage,
2001; Wolf, 2003). In particular, the release of cortisol (or its
administration in placebo-controlled pharmacological studies) is
associated with impaired memory retrieval (deq Uervain et al.,
2000; Kuhlmann, Piel, & Wolf, 2005; Wolf et al., 2001). In a
sample of Special Forces candidates evaluated in the course of an
intense naturalistic stressor, high levels of cortisol secretion were
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2HOPE ET AL.
associated with impaired cognitive performance relative to a con-
trol group (Taverniers, Taylor, & Smeets, 2013).
A second problem with “attentional narrowing” accounts, as
noted by Davis and Loftus (2009), is that emotional arousal may
not narrow attention in all cases, and certainly not to predictable
stimuli. In fact, stress impairs executive function (Schoofs, Wolf,
& Smeets, 2009), including the ability to control attention or where
it is directed (Banks, Tartar, & Welhaf, 2014). Laboratory work
demonstrating supposed attentional narrowing typically uses rela-
tively uncomplicated or unambiguous stimuli (e.g., the deliberate
presentation of a single weapon in much of the so-called weapon
focus literature) in a third-hand presentation format (e.g., video or
slides). In real life situations, there are likely to be many conflict-
ing draws on attention as a function of (a) a more complicated
interactive scene and (b) the need to respond to an incident, which
may further deplete cognitive resources. Taken together, these
factors make it difficult to predict what will be remembered from
stressful naturalistic events. Thus, the first aim of the current study
was to examine the effects of arousal, experienced in a challenging
yet controlled event, on officer recall as a function of response
role.
Attentional Load, Training, and Expectations
Both operationally active and lay (or civilian) witnesses’ mem-
ories for an incident may be impaired as a consequence of high
stress levels during encoding. However, active witnesses have the
additional task of deriving an appropriate response strategy (in
light of various and potentially transient contextual factors, and
taking into account their own safety and that of others in the
vicinity), planning the effective execution of that strategy and then
taking action accordingly (see Eyre & Alison, 2007). Obviously,
an important contextual factor in minimizing the cognitive “drain”
of such activities should be the effective training of operationally
active witnesses such that response options are fluently available
(e.g., recognition-primed decision-making, Klein, 1998; see also
Clark, 2008; Healy, Kole, & Bourne, 2014). However, by its
nature, training might set expectations about the likely outcomes of
particular scenarios. According to Neisser’s (1976) notion of the
“perceptual cycle,” individuals perceive, interpret, and revise their
understanding of information using both top-down and bottom-up
processes. Certain stimulus properties elicit an attentional response
and, inevitably, expectations derived from existing schemas guide
further interpretation and recollection (e.g., Tuckey & Brewer,
2003; see also Most, Scholl, Clifford, & Simons, 2005). For
example, when specific schemas are activated, visual attention is
likely to be directed toward schema-relevant items (Eberhardt,
Goff, Purdie, & Davies, 2004). Evaluating an emerging situation
and responding appropriately are critical police activities, partic-
ularly for high stakes incidents. However, these activities will
likely take place under conditions of time pressure, depleted cog-
nitive resources and additional cognitive load—conditions under
which such evaluations, and subsequent recollection, may be par-
ticularly vulnerable to schema-driven or expectation-based errors
(e.g., Kleider, Pezdek, Goldinger, & Kirk, 2008; Sherman &
Bessenoff, 1999; Tuckey & Brewer, 2003; see Betts & Hinsz,
2013). If, under such circumstances, the available schema posits
that Action X is usually followed by Action Y, then “interpreta-
tions or classifications made on the basis of emotion- or
expectation-weakened identification criteria, unchecked by discon-
firming evidence, enter long-term memory uncorrected and be-
come the basis of distorted witness reports” (Davis & Loftus,
2009, p. 182). Thus, the second aim of the current study was to
examine whether active witnesses, like other professionals in high
reliability roles (e.g., Plant & Stanton, 2012), may be vulnerable to
expectation-based errors that impact the accuracy of their subse-
quent accounts.
Current Study
Our research was motivated by two main research questions.
First, we examined whether there were differences in the accounts
provided by operational (active) and nonoperational (observer)
witnesses where both had been exposed to the same incident. To
achieve this we put pairs of police officers into an immersive and
stressful simulated scenario where one officer was instructed to
respond as they would usually in the course of their duty (i.e., an
active operational witness) while the other was instructed to sim-
ply observe the scenario. To our knowledge, this is the first study
to adopt a methodology of this kind to examine the effects of
active role involvement in an eyewitness context. We predicted
that officers allocated to the active response role would show an
increased physiological response in the scenario, reflecting by
proxy, increased stress as a function of their active response role.
On the grounds of this increased arousal, and consistent with the
theoretical accounts outlined above, we predicted that the quality
and quantity of free recall reported by the active response officers
would be impaired relative to the observers.
In addition to the free-recall task, both witnesses were asked a
series of detailed closed questions. These questions were included
to contribute data and inform current practice and policy in the
investigation of shooting incidents. In a number of recent cases,
after providing written statements, operational witnesses have been
asked to respond to long lists of additional, detailed closed ques-
tions (see Dodd & Travis, 2014). Research on investigative inter-
viewing has long documented the problems associated with closed
question interviewing approaches (cf. open questions), not least
the dangers of leading, closed questions (e.g., Sharman & Powell,
2012). Current “gold standard” interview techniques, such as the
Cognitive Interview (Fisher & Geiselman, 1992) and the National
Institute for Child Health and Development Protocol (NICHD;
Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007) endorse
the use of open-ended questions with compliant witnesses (for
overview, see Vrij, Hope, & Fisher, 2014). In the current study,
participants were exposed to an extended set of closed questions
modeled on the question style often adopted by investigators (see
Method). In addition to documenting accuracy rates for such
questions, we were particularly interested in the recall performance
of active witnesses (relative to their observing cowitnesses) on
questions pertaining to use and location of the target’s weapon as
such questions are, unsurprisingly, a central focus of investigations
following police shootings. Specifically, we predicted that addi-
tional demands on the resources of operational witnesses in con-
junction with higher arousal and stress levels, both likely to occur
when police officers were required to discharge (or consider dis-
charging) their own weapon, would impair recall for such infor-
mation.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
3
MEMORY AND THE OPERATIONAL WITNESS
Our second independent research question concerned the possi-
ble effect of expectations as they relate to memory performance
under challenging operational conditions. Deliberately, the sce-
nario was designed to trigger schema-driven expectations regard-
ing the likely action of the perpetrator. In the final sequence of the
scenario, the perpetrator who was at this point known to be armed,
turned quickly to face the officers, throwing out his hands in front
of him. However, the weapon (a gun) remained in the waistband of
his trousers. In light of the memory deficits reported in previous
research (e.g., Morgan et al., 2013) and well-documented effects of
schema-reliance (e.g., Kruglanski & Freund, 1983; Tuckey &
Brewer, 2003), we predicted that memory reports provided by
officers in the Active (cf. Observer) condition may be particularly
vulnerable to the expectation-based error that the perpetrator
would point the weapon at them. For the current study, it should be
noted that our research questions were entirely focused on poste-
vent recollection by officers and not behavioral outcomes, such as
shoot/no-shoot decisions that are explored elsewhere (e.g., Akinola
& Mendes, 2012; Nieuwenhuys, Savelsbergh, & Oudejans, 2012).
Method
Participants
Eighty-seven serving Canadian law enforcement officers affili-
ated to a metropolitan force were recruited. Because of technical
difficulties, incomplete data or single participant sessions (due to
unavoidable no-shows), the final sample comprised 76 partici-
pants. Participants (64 males) were aged 22 to 59 years of age
(M⫽37 years, SD ⫽7.99). Most of the sample was at Constable
rank (94%) with the remainder at Detective (3%) and Sergeant
(3%) rank. Recruited officers represented a range of experience
with an average length of service of 147.75 months of service
(M⫽12.31 years, SD ⫽89.47 months).
The purpose or nature of the study was not revealed in ad-
vance—participants were led to believe that they were taking part
in research related to officer response. Officer participation in the
research was voluntary and took place during work hours at
prearranged times with the full agreement of shift supervisors.
Although the research was organized in collaboration with the
training division of the force, test sessions did not constitute formal
training events. Officers were not paid for their participation and
received no work-related rewards for taking part. In addition to
adhering to standard ethical principles and considerations, detailed
Informed Consent procedures assured officers that their individual
professional performance was not being assessed and reiterated
confidentiality procedures.
Design
Officers were randomly assigned to either the “Active Officer”
or “Observer Officer” condition during the encoding phase. Con-
ditions were paired such that each Active Officer viewed the
scenario with an Observer Officer. All participants completed the
same test materials. The experimental data were collected over a
5-day period. In a typical test day, eight pairs took part in live
scenarios.
Materials
Briefing video. Before deployment into the main scenario, all
officers viewed a short “briefing” video that depicted the initial
hostage-taking incident. It showed students taking part in a
classroom-based seminar with their professor and was filmed on a
cell phone (as if from the perspective of one of the students). The
perpetrator (an apparently disgruntled student) entered the class-
room and engaged the professor in a discussion about poor grades.
The perpetrator became increasingly agitated, drew a knife from
his pocket, and took the professor and a student hostage. Toward
the end of the video, the other students are seen rushing from the
room, initiating calls to the police. The film, recorded using an
iPhone, lasted 2 min 10 sec and was of high quality with clear
audio.
Experimental scenario. The scenario was developed in the
course of extensive discussion with firearms instructors and police
trainers. Three key objectives guided the scenario development:
the need for (a) a relatively complex scenario which the partici-
pants could be questioned about in detail, (b) a scenario with
challenging elements that would produce a natural physiological
stress response, and (c) a realistic scenario officers might encoun-
ter in the course of their duty and be reasonably expected to
respond to.
All officers encountered the same “augmented reality” scenario,
lasting 4 min, which incorporated prerecorded and live elements.
The prerecorded elements of the incident were presented as “live”
CCTV footage and were integrated with fully scripted, tightly
controlled live elements, reenacted for each pair of participants,
using three actors (one male “perpetrator,” two male “hostages”).
At the outset, the perpetrator was shown via a CCTV-feed, threat-
ening the two hostages in a hallway (outside the classroom where
they had originally been located in the briefing video). One of the
hostages was then released and could be seen walking down a
hallway, appearing first in the “CCTV footage” and then in reality
through a window in the classroom through to the same hallway.
This visual device was incorporated to fully establish the link
between the apparent CCTV footage and elements of the scenario
enacted in real-time activities. This method appeared successful as,
during debriefing, a number of officers indicated surprise that any
elements had been prerecorded and all officers reported the sce-
nario as a single integrated event. Shifting from prerecorded foot-
age to live interaction, the perpetrator then entered the classroom
(where the officer participants were located) using a hostage as a
shield and holding a knife to the hostage’s neck. After issuing
various demands, he set the hostage free and threw the knife to the
ground before retreating to the hallway and closing the door to the
classroom. He could then be seen, on the CCTV, tucking the gun
into the waistband of his jeans. In the final live interaction, the
perpetrator reentered the classroom in an agitated manner. The gun
remained in the waistband of his trousers throughout.
Recall tasks. The recall tasks comprised two different re-
sponse formats—Free Recall and Closed Questions. In the Free
Recall task, which was presented first, there were two sections.
The first section requested details of the briefing information
encountered at the outset (Briefing Phase). The second section
requested details of the main scenario involving the live CCTV
footage and perpetrator (Response Phase). Instructions at the start
of both sections asked participants to report as much information
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
4HOPE ET AL.
as they could remember about the event, emphasizing that their
account should be “as complete and accurate as possible.” Partic-
ipants were also instructed against guessing. In the Closed Ques-
tions task, participants answered 94 questions adapted, in terms of
style and content, from the type of questions posed by external
investigators in such circumstances (as mentioned by Dodd &
Travis, 2014). Ninety-one of these questions sought factual and
verifiable information about the incident, three questions asked for
a more subjective assessment of the perpetrator (e.g., “What was
the demeanor of the perpetrator?” and “Describe his facial expres-
sion”). A subset of these closed questions (14 questions; target
questions) were identified by legal and police training advisors as
important questions from an investigative perspective with respect
to an officer’s recall of the critical response phase. These question
subsets were categorized as “Perpetrator Action” (e.g., Did he turn
to the left or to the right? What position did he move his arms to?),
“Officer Response” (e.g., What action(s) did the other officer(s) in
the room take at this point? How many shots were fired and by
whom?), and “Perpetrator Weapon” (e.g., Did the culprit dis-
charge a weapon during the incident? If yes, how many shots did
he fire? Where was the gun at the end of the scenario?).
Procedure
Participants were recruited in pairs and were randomly allocated
to the role of Active Officer or Observer Officer by virtue of their
choice of seat in the waiting area. Both officers were fitted with
Polar Heart Rate monitoring belts, equipped with safety glasses,
and given general instructions about their role. Active Officers
were instructed to respond to the scenario event as they would
normally in the course of their duty. Observer Officers were
instructed to take no active response role and to simply observe
what happened during the scenario. A verbal briefing by a “senior
officer” informed them that there was an ongoing hostage situation
involving a male perpetrator armed with a knife in a remote
corridor area. They were told that negotiations had been underway
but that the perpetrator had stopped communicating in the past 30
min. The Active Officer was provided with a training handgun
loaded with five blank rounds (i.e., their weapon was available for
discharge) and informed s/he was part of an initial response team
with the objective of moving forward into the classroom adjacent
to the corridor containing the hostages to tactically assess the
situation and intervene, or advise other teams available to inter-
vene as necessary.
Before entering the critical response phase, both participants,
seated side by side, viewed the briefing video on a laptop screen.
They were told that this was “cell phone footage that a witness
captured as the situation developed this morning.” This concluded
the briefing phase. At the outset of the critical response phase, both
officers were then taken into the classroom where they could view
the live CCTV footage of the ongoing incident. Both were
equipped with headsets to ensure they could hear the footage
soundtrack and also to mitigate any effects of any verbalization by
their coparticipant (some participants initiated verbal comments
with respect to the evolving scenario; we noted no apparent effect
of this verbalization on reporting and observed that as the scenario
escalated, verbal commentary typically ceased). A confederate
officer initiated the footage when they entered the room and
ensured both participants stood in preallocated side-by-side con-
tainment positions (behind a desk) that shared an equally clear
view of both the CCTV footage (within 2 feet) and live action
space (within 15 feet). A senior police instructor monitored each
trial from a health and safety perspective and ended each trial with
a whistle blast. Each trial was captured on two digital cameras.
There was a delay of 45–50 min between the end of the response
phase and the recall tasks. The purpose of this delay was twofold.
First, to allow attenuation of any immediate stress response and
second, to mimic a minimum delay before an initial interview with
an investigator. During this time, officers were seated at separate
desks in silence. No discussion with their partner (or anyone else)
was permitted and this instruction was closely monitored by the
researcher. After the delay, participants completed the recall tasks.
When they had completed the free recall component, they were
given the closed questions and worked through them in sequential
order. Drawing on an approach devised by Scoboria and col-
leagues (e.g., Scoboria & Fisico, 2013) to better understand the
meaning of “Do Not Know” responses in interview contexts,
participants were then asked to clarify any “Do Not Know”/
“Unknown” responses by selecting one of four options for each
response of that type: (a) I did not report an answer because the
information was not present in the event (no one could answer this
question); (b) I did not report an answer because I could not recall
the specific information from the event (someone else might be
able to answer this question, but I cannot); (c) I did not report an
answer because I truly do not remember (I do not know whether it
is possible to answer this question or not from the information in
the event); or (d) I did not report an answer because I was not
confident enough that it was correct (I could provide an answer if
pushed, but it might be wrong).
No time limits were imposed and participants took, on average,
2 hr to complete the recall tasks and final classification task.
Officers were then fully debriefed and thanked for their contribu-
tion to the research.
Coding
Both free recall sections (Briefing Phase and Response Phase)
were coded for quantity and accuracy. Using a coding scheme
adapted from Hope, Gabbert, and Fraser (2013), each unit of
information reported was categorized as either correct or incorrect.
A second coder, blind to experimental condition, coded a random
sample (15%) of the free recall reports. Intercoder reliability for
briefing phase recall was ⫽ .72, p⬍.001, 95% confidence
interval (CI) [0.47, 0.98] and for response phase recall was ⫽
.81, p⬍.001, 95% CI [0.58, 1.00], suggesting substantial to high
levels of agreement (Landis & Koch, 1977). Accuracy rate was
calculated by dividing the total correct items by total responses
(correct and incorrect).
For the 94 closed questions, model correct answers were agreed
by the research team in a detailed review of the scenario videos.
No divergence occurred during any trial that would have resulted
in a different possible answer for any question. The closed ques-
tion responses were coded as either correct or incorrect. “I do not
know” (and variations thereof e.g., “unknown,” “not sure”) re-
sponses were also recorded. Responses to questions concerning the
demeanor of the perpetrator elicited subjective responses (“he
looked angry”), and consistent with coding conventions in previ-
ous research, were not included in the analysis of recall data.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
5
MEMORY AND THE OPERATIONAL WITNESS
Results
Individual officers took part in pairs (in this case, distinguish-
able pairs; Kenny, Kashy, & Cook, 2006) such that all participants
were exposed to the same highly controlled scripted and videoed
stimuli. Given this dyadic design structure, it was necessary to
assess the degree of nonindependence before proceeding with the
main analyses. Following Kenny et al.’s (2006) approach, we
conducted a preliminary analysis to establish the extent to which
responses within the pairs were correlated. First, the dataset was
restructured to apply SPSS (Chicago, IL) syntax, developed by
Alferes and Kenny (2009, see Supplementary Materials), to com-
pute the Pearson product–moment correlation coefficient between
dyad members and perform a ttest of the null hypothesis that the
population correlation is 0. For the free recall data for the 38 intact
pairs in the dataset, r(38) ⫽.12 [⫺.21, .42], t(36) ⫽.74, p⫽.47.
The same analysis was conducted for the cued recall data. The
correlation and associated ttest were also nonsignificant; r(38) ⫽
.21 [⫺.12, .50], t(36) ⫽1.28, p⫽.21. Given independence
between dyad member scores for both recall tasks, each indi-
vidual was used as the unit of analysis in subsequent analyses;
see Kenny et al. (2006) for informative discussion around
measuring (non)independence in dyadic data.
Physiological Response
Usable heart rate recordings were available for 61 participants in
the sample (33 participants in the Active Officer condition and 28
participants in the Observer Officer condition). The mean resting
heart rate (HR) recorded over a 10 min period 1 hr after the critical
response phase was 68.01 beats per minute (bpm; SD ⫽9.17) and
was roughly equivalent between groups (Active Officers M⫽
68.81, SD ⫽8.68; Observer Officers M⫽67.13, SD ⫽9.80). HR
measurements recorded during the response phase showed a range
of physiological arousal response with a range of 76–164 bpm (see
Table 1). The average maximum HR (mHR) recorded during the
response phase were significantly higher for Active Officers than
Observer Officers, t(59) ⫽2.89, p⫽.005, d⫽0.75, 95% CI [0.22,
1.28]. HR variability (HRV; Thayer, Ahs, Fredrikson, Sollers, &
Wager, 2012), another measure of workload under stress measured
over a 1 min period during the response phase, also differed
between groups, t(59) ⫽ ⫺2.30, p⫽.025, d⫽0.59, 95% CI [0.06,
1.11]. Lower HRV is associated with increased stress (see meta-
analysis by Thayer et al., 2012).
Recall Performance
Free recall. For the free recall data, the dependent variables of
interest were quality (as reflected in the accuracy of accounts) and
quantity (as reflected in the amount of information provided). As
quantity is most usefully examined in terms of the amount of
correct and incorrect items, it should be noted that the quality and
quantity measures are not independent.
Recall of briefing phase. There was no significant difference
between conditions for the amount of correct, t(74) ⫽0.06, p⫽
.96, d⫽0.01, 95% CI [⫺0.44, 0.47] or incorrect, t(74) ⫽0.15,
p⫽.88, d⫽0.03, 95% CI [⫺0.42, 0.49] information reported
about the events viewed in the briefing video. Similarly, there was
no difference between conditions in the overall accuracy rate for
this information, t(74) ⫽ ⫺0.35, p⫽.73, 95% CI [⫺0.04, 0.02],
d⫽0.13, 95% CI [⫺0.58, 0.32]; see Table 2.
Recall of response phase. There was a difference between
conditions for the number of correct details reported about the
critical response phase, such that Active Officers reported signif-
icantly fewer correct details than Observer Officers (see Table 2),
t(74) ⫽ ⫺2.74, p⫽.008, d⫽0.63, 95% CI [0.15, 1.09]. There
was no difference in the amount of incorrect information reported
between the experimental groups, t(74) ⫽0.87, p⫽.39, d⫽0.20,
95% CI [⫺0.25, 0.66]. The overall accuracy rate did not differ
between the groups either, t(61) ⫽ ⫺1.46, p⫽.15, d⫽0.32, 95%
CI [⫺0.77, 0.14].
Integrating Physiological Response and Free
Recall Performance
The analyses suggest a link (a) between the role the officer was
assigned to (active vs. observer) and the degree of arousal expe-
rienced (using maximum HR as a proxy measure for the peak of
that arousal) and (b) between role and free recall performance,
specifically the amount of correct information reported about the
response phase. These associations might be formulated as X¡M
and X¡Y, respectively, in terms of the meditational model X¡
M¡Y(where Xis the independent variable, Yis the dependent
variable and Mis the mediating variable; Baron & Kenny, 1986).
In line with Baron and Kenny’s (1986) recommendations for
establishing mediation, we constructed three regression equations.
First, regressing mHR on experimental condition (X¡M) was
statistically significant, ⫽⫺15.72, 95% CI [⫺26.77, ⫺4.62],
t(59) ⫽ ⫺2.89, p⫽.005. Second, regressing correct response
phase free recall performance on experimental condition (X¡Y)
was significant,  ⫽ 7.17, 95% CI [2.17, 12.34], t(74) ⫽2.74, p⫽
.008. Finally, regressing free recall performance on both experi-
mental condition and mHR rendered mHR significant (⫽⫺.18,
95% CI [⫺.31, ⫺.06], t(58) ⫽ ⫺2.76, p⫽.008) but not experi-
mental condition,  ⫽ 4.81, 95% CI [⫺1.06, 10.33],
t(58) ⫽ ⫺1.69, p⫽.09. The indirect effect of the independent
variable (role) on the dependent variable (correct free recall) via
Table 1
Means, SDs, and 95% Confidence Intervals (CIs) for Maximum Heart Rate (MHR) and Heart
Rate Variability (HRV) by Experimental Group
Active Officers Observer Officers
Mean (SD) 95% CI Mean (SD) 95% CI
Max HR (mHR)
ⴱⴱ
126 (19.95) [119.38, 132.89] 110 (22.58) [102.44, 118.94]
HR variability (HRV)
ⴱ
22.34 (14.89) [17.81, 27.51] 32.83 (20.63) [25.58, 40.35]
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
6HOPE ET AL.
the mediator was significant, Sobel Test ⫽2.08, p⫽.02. Baron
and Kenny (1986) state that a variable Mfunctions as a mediator
when the significant effect of Xis rendered nonsignificant after
controlling for M. Thus, in the current analyses, officer role was
related to the degree of arousal experienced during the critical
response phase and the effect of role on correct information
recalled was mediated by arousal.
Although regressing HRV on experimental condition was sta-
tistically significant,  ⫽ 10.49, 95% CI [1.41, 19.11],
t(59) ⫽ ⫺2.30, p⫽.025, HRV did not mediate the effect of role
on recall. Experimental condition continued to predict free recall
performance ( ⫽ 7.65, 95% CI [1.52, 14.06], t(58) ⫽2.59, p⫽
.012, but HRV did not, ⫽⫺.007, 95% CI [⫺.17, .15],
t(58) ⫽ ⫺0.09, p⫽.93 (see Discussion).
Closed Questions
The overall accuracy rate for the closed questions was 57%
(SD ⫽11.25) with accuracy ranging from 28% to 76% and no
difference between conditions although there was a trend to greater
accuracy, in terms of overall accuracy rate, in the observer condi-
tion (Active M⫽.55, SD ⫽.13, 95% CI [.51, .59]; Observer M⫽
.60, SD ⫽.08, 95% CI [.57, .62]), t(73) ⫽ ⫺1.78, p⫽.08, d⫽
0.46, 95% CI [⫺0.002, 0.92]. On average, officers wrote a Do Not
Know response for 17% of the closed questions with no difference
between experimental groups in the mean frequency of Do Not
Know responding (Active M⫽.19, SD ⫽.11, 95% CI [.16, .23];
Observer M⫽.15, SD ⫽.08, 95% CI [.12, .18]), t(73) ⫽1.62, p⫽
.11, d⫽0.42, 95% CI [⫺0.05, 0.88]. With respect to Do Not
Know responses, participants were asked to categorize such re-
sponses to one of four categories. Notably, there was a significant
effect of Officer Role on the frequency of selection of one of the
categories (Category A: “I did not report an answer because the
information was not present in the event;” see Table 3).
Target questions. Recall that the target questions were a
subset of the closed questions identified by legal and police train-
ing advisors as critical from an investigative perspective. For
analysis, the average number of correct, incorrect and Do Not
Know responses were calculated for each of the three subsets,
Precursor Perpetrator Action, Officer Response, and Perpetrator
Weapon. For Perpetrator Action and Officer Response, there were
no significant differences between conditions for mean number of
correct, incorrect or DK responses (see Table 4).
With respect to questions pertaining to the perpetrator’s weapon,
there was no difference between conditions for the number of ques-
tions answered correctly or with a Do Not Know response. However,
there was a significant difference between conditions for incorrect
responses (see Table 4) and accuracy rate (Active M⫽.55, SD ⫽.33,
95% CI [.44, .66]; Observer M⫽.71, SD ⫽.31, 95% CI [.60, .81]),
t(73) ⫽2.13, p⫽.04, d⫽0.50, 95% CI [0.03, 0.96].
Memory for Weapon-Related Details
Of the 39 participants placed in the role of the Active Officer, 33
of them discharged their weapon during the scenario (85%). We
conducted an additional coding of the free recall data to examine how
Table 2
Means, SDs, and 95% Confidence Intervals (CIs) for Correct and Incorrect Items Reported and
Accuracy Rate by Experimental Group
Active Officers Observer Officers
Mean (SD) 95% CI Mean (SD) 95% CI
Briefing phase
Correct 47.46 (17.37) [41.97, 52.68] 47.24 (16.01) [42.32, 52.45]
Incorrect 4.72 (5.66) [3.16, 6.86] 4.54 (4.21) [3.25, 6.00]
Accuracy rate .91 (.08) [.88, .94] .92 (.07) [.90, .94]
Critical response phase
Correct
ⴱⴱ
38.67 (10.17) [35.61, 41.97] 45.84 (12.59) [42.02, 50.03]
Incorrect 1.64 (2.18) [1.00, 2.40] 1.29 (1.05) [.97, 1.66]
Accuracy .96 (.04) [.95, .97] .97 (.02) [.96, .98]
ⴱⴱ
p⬍.01.
Table 3
Mean Proportion of Do Not Know (DK) Responses Allocated to Each Category by Experimental Group
DK response
Active Observer
tstatistic
a
d95% CIM(SD) 95% CI M(SD) 95% CI
Did not report an answer because the information was not present (A)
ⴱ
.17 (.24) [.09, .27] .05 (.16) [.01, .12] 2.22 (p⫽.03) .59 [.05, 1.12]
Did not report an answer because I could not recall the specific
information (B) .55 (.32) [.44, .67] .57 (.39) [.43, .70] ⫺.22 (p⫽.83) .05 [⫺.58, .47]
Did not report an answer because I truly do not remember (C) .18 (.21) [.11, .26] .28 (.34) [.16, .42] ⫺1.34 (p⫽.19) .36 [⫺.88, .17]
Did not report an answer because I was not confident that it was be
correct (D) .09 (.15) [.04, .15] .09 (.19) [.04, .17] ⫺.005 (p⫽.99) .00 [⫺.52, .52]
a
df ⫽56 for a between-subjects t-test. CI ⫽confidence interval.
ⴱ
p⬍.05.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
7
MEMORY AND THE OPERATIONAL WITNESS
many participants spontaneously reported that the perpetrator pointed
a weapon at them/in their direction in response phase (note that the
gun had remained in the perpetrators waistband throughout that
scene). Overall, 18% of participants spontaneously reported that a gun
was pointed at them in the final part of the scenario. There was no
association between Officer Role and likelihood of reporting the
perpetrator pointing a gun toward the officers in the final scenario:
15% of Active officers and 22% of Observer officers reported seeing
a gun in the hands of the perpetrator,
2
⬍1.
Discussion
Operationally active witnesses did not differ from their observer
counterparts with respect to their recall of the briefing phase
encountered before immersion in the critical response phase. Re-
call of the initial briefing phase arguably represents a baseline
recall measure as arousal levels would have been equivalent at
time of encoding (and our baseline HR measures suggest there are
no other systematic physiological differences between the groups).
However, an interesting difference emerged between active and
observer officers in their recall of the critical response phase.
Operationally active witnesses reported significantly fewer correct
details about the scenario than observer witnesses. However, there
were no differences, according to role type, in the overall accuracy
rate of information reported. Thus, operational witnesses, in free-
recall tasks at least, were able to sustain the accuracy of their
accounts. Indeed, accuracy rates for freely reported information
were very high across both operational and nonoperational wit-
nesses (ⱖ92%). In light of the high profile political, legal, and
investigative contexts such accounts are evaluated in, the adoption
of a conservative reporting strategy is not particularly surprising,
and would explain the high levels of accuracy observed here (see
Hope et al., 2013 for further discussion of this issue). Nonetheless,
that operationally active witnesses reported significantly fewer
correct details about the critical response phase than their nonop-
erational observer counterparts is important.
Our physiological data may help account for this finding. Op-
erationally active witnesses showed significantly higher levels of
physiological arousal, as marked by higher HRs and lower HRV,
during the critical response phase of the scenario in comparison
with their nonoperational counterparts. It is noteworthy that sig-
nificantly different HR measures were recorded for active wit-
nesses despite the fact that officers in both active and observer
roles were exposed to the same critical scenario and stood side by
side while the scenario unfolded. In other words, higher HRs and
lower HRV did not reflect increased physical activity—in fact, we
deliberately limited the potential for differential physical move-
ment through the enforcement of predetermined containment po-
sitions. Therefore, active witnesses in the current study experi-
enced higher levels of physiological arousal or stress response as
a function of the demands of their operational response role.
Mediational analyses revealed that the observed effects of role
on free recall performance were related to level of stress, as
indexed by the maximum HR recorded during the critical response
phase. Thus, while it is important to consider the role of a witness
to an event, the degree of arousal experienced is also an important
factor. Although HR variability recorded during the critical re-
sponse phase reflected relatively increased stress workload for
active witnesses, a similar meditational relationship was not ob-
served. This result can most likely be accounted for by the sam-
pling period for both measurements. HR typically peaked at the
critical final moments in the scenario, in the same 2–3 sec, when
the perpetrator turned toward the officers (i.e., the point at which
85% of active officer discharged their weapon). As such, the
maximum HR measure reflects a specific instantaneous period of
elevated stress. Conversely, HRV was calculated, in line with
conventions over a longer period during the response phase (1 min;
see Spierer, Griffiths, & Sterland, 2009). As such we would not
necessarily expect HRV to mediate recall performance across this
longer period. Nonetheless, it is noteworthy that this measure
reflected increased stress for the active witnesses across the critical
response phase.
Although high accuracy rates were obtained for both groups in
these free recall reports, requiring officers to respond to closed
questions produced a very different pattern of results. Accuracy
rates for closed questions were comparatively low for both Active
and Observer Officers—the average accuracy rate was 57% mean-
Table 4
Mean Number of Target Questions Answered Correctly, Incorrectly, and Using A Do Not Know Response by Participant Role and
Question Category
Target questions
Active Observer
tstatistic
a
d95% CIM(SD) 95% CI M(SD) 95% CI
Perpetrator action
Correct 3.94 (2.11) [3.27, 4.58] 4.43 (1.92) [3.75., 5.03] ⫺1.04 (p⫽.30) .24 [.21, .70]
Incorrect 3.49 (2.11) [2.84, 4.12] 2.70 (1.80) [2.11, 3.29] 1.62 (p⫽.11) .40 [⫺.5, .86]
Do not know 2.61 (2.04) [2.00, 3.25] 2.2 (1.63) [1.69, 2.77] .94 (p⫽.35) .22 [⫺.23, .68]
Officer response
Correct 2.51 (1.09) [2.16, 2.85] 2.37 (1.86) [2.00, 2.736] .51 (p⫽.61) .10 [⫺.36, .55]
Incorrect .72 (.82) [.45, .97] .70 (.70) [.48, .91] .09 (p⫽.93) .02 [⫺.43, .48]
Do not know .74 (.91) [.49, 1.03] .54 (.90) [.28, .83] .98 (p⫽.33) .22 [⫺.23, .67]
Weapon position
Correct 2.21 (1.30) [1.83, 2.58] 2.73 (1.30) [2.26, 3.13] 1.75 (p⫽.08) .41 [.06, .86]
Incorrect
ⴱ
1.15 (1.08) [.82, 1.51] .65 (.88) [.37, .95] 2.21 (p⫽.03) .51 [.04, .97]
Do not know .64 (1.01) [.36, .97] .48 (.90) [.23, .78] .70 (p⫽.48) .16 [⫺.29, .62]
a
df ⫽74 for a between-subjects t-test. CI ⫽confidence interval.
ⴱ
p⬍.05.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
8HOPE ET AL.
ing that just under half of the questions were answered incorrectly
or with a Do Not Know response. More important, there was a
difference in response accuracy for information pertaining to the
perpetrator’s weapon (i.e., a legally relevant subset of questions)
with active officers significantly more likely to provide incorrect
information than observer officers.
These are important findings and, to our knowledge, this is the
first study to document a physiological difference between wit-
nesses who have different roles in responding to the same incident,
observe that the effect of role operates through an arousal mech-
anism and demonstrate differences in memory performance for
operational versus nonoperational witnesses. As such, these find-
ings confirm the merits of considering the role of “operational
witness” when evaluating their statements.
Despite showing a physiological profile consistent with an in-
creased stress response, active officers in the current study clearly
did not experience a generalized “catastrophic failure” of memory
(in terms of the overall quantity of information reported). How-
ever, consistent with previous findings (e.g., Hope et al., 2013)
they did report significantly less correct information than nonop-
erational observer counterparts. Examination of the errors made by
active witnesses in response to closed questioning highlights par-
ticular areas of vulnerability in their recollection of the incident.
Specifically, as illustrated by performance on target questions, the
recall performance of active witnesses was significantly impaired,
relative to observer officers, for critical information about the
weapon in the final moments of their interaction with the perpe-
trator (i.e., when threat level was greatest). Active witnesses re-
ported less information than observers in response to questions
about the weapon and their responses were less accurate. Active
officers were also more likely than their observer counterparts to
categorize their use of a Do Not Know response as “I did not
report an answer because the information was not present in the
event”. This suggests that details of the event were either not
encoded in the first place or were no longer accessible. At first
glance, the findings of the current study appear to be inconsistent
with classic “attentional narrowing” accounts that propose a recall
advantage for central or important stimulus information over pe-
ripheral information (e.g., Safer, Christianson, Autry, & Österlund,
1998). However, this would be a premature and likely inaccurate
conclusion—particularly in the absence of data for a control,
nonarousing version of the scenario. Observer officers in the
current study were exposed to the same arousing encounter and
displayed elevated HRs during the critical scenario (averaging 112
bmp) contrasting with their baseline HRs (averaging 67 bpm).
However, this group achieved reasonably high accuracy rates for
questions pertaining to the weapon, almost certainly a high priority
stimulus and focus of attention during the critical response phase.
There are a few potential explanations. First, it may be that
observers simply did not meet an arousal threshold likely to impair
recall. It is worth noting that the average HR for observers is close
to the 110 bpm threshold beyond which the sympathetic nervous
system (SNS) is triggered (cf. the average rate of 126 bpm for
active witnesses that is well beyond this threshold; see Woody &
Szechtman, 2011). Therefore, the level of arousal experienced by
observers, in the absence of further competing demands, did not
necessarily impede the processing of important information in the
scene and, thus, the findings are likely to be consistent with
previous literature. As such, moderate arousal as a function of
merely witnessing a threatening incident was not responsible for
reduced recall performance of active witnesses (relative to observ-
ers) in the current study. Future research might consider innovative
methodological approaches where the meaning of the important
stimuli is altered depending on environmental context (e.g., Pickel,
1999) and arousal level to explore these comparisons further for
operationally active and nonoperational observer witnesses.
Operationally active witnesses did not show a recall enhance-
ment for critical stimuli—in fact, relative to observer witnesses,
they reported fewer correct details about the critical response
phase and had poorer accuracy for questions about the weapon.
There are a number of possibilities as to why this might be the
case. First, in light of the physiological results, it is possible that
active witnesses experienced greater stress responses than their
observer counterparts that may have contributed to memory im-
pairment for details of this final phase of the scenario where the
threat level (and likely associated stress) was highest. This pattern
of results is consistent with the inverted U-shaped curve predicted
by arousal theories to account for performance decrements when
moderate stress levels are exceeded (Morley & Farr, 2012). Sec-
ond, additional cognitive load as a function of response role may
have impaired their ability to process information about details of
the final scene. This deficit is particularly evident for information
pertaining to the fate of the perpetrator’s weapon. Given that
operational demands were likely to be at their greatest at this
moment (i.e., attempts to attenuate immediate lethal threat), these
findings are unsurprising and likely reflect reduced information
processing resources. Third, the interaction between heightened
stress and additional cognitive load associated with responding and
attempting to neutralize the threat, may have produced impaired
encoding (and/or retention) of details of the scene. As such, our
findings are consistent with Morgan and Southwick (2014) who
have argued on the basis of data obtained in challenging military
training contexts, that memory for stressful events can be vulner-
able to error and suggestion—contrary to predictions that emo-
tionally arousing events will be remembered better than neutral
events (see also Engelhard, van den Hout, & McNally, 2008;
Lommen, Engelhard, & van den Hout, 2013; Morgan et al., 2013).
In the current study, the increased stress experienced by active
witnesses may have led to more generalized processing and re-
sulted in the rapid extraction of gist information (e.g., Payne et al.,
2002; see also Qin, Hermans, van Marle, & Ferna´
ndez, 2012) that
they were able to report accurately in a free recall account, but that
was not sufficiently specific to produce accurate detailed informa-
tion about the weapon when probed by closed questions.
The current data might also be considered in light of compen-
satory control models (e.g., Hockey, 1997). When processing
resources are compromised (e.g., because of threat), individuals
make strategic adjustments in the allocation of resources to main-
tain performance on high priority task goals (Hockey, 1997). Such
adjustments often produce decrements on secondary tasks or am-
plify trade-offs (Hockey, 1993; Hockey & Hamilton, 1983). It has
typically been assumed, here and in other research, that details
associated with a deadly weapon constitute the critical stimuli.
However, there are important contextual factors to take into ac-
count. First, training guidance on firearms practice (e.g., Armed
Policing Authorised Professional Practice, 2013) often recom-
mends aiming to strike central body mass depending on the type of
weapon being used and training received, which means that rather
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
9
MEMORY AND THE OPERATIONAL WITNESS
than sustaining their attention on a weapon they already know to
be present, officers will refocus their attention on a target’s critical
mass. Second, it may well be the case that once an officer has
committed to a decision to discharge their weapon, other compet-
ing priorities take precedence including monitoring the immediate
environment for further risk factors, securing the safety of oneself
and others and planning the next action. Previous research has
documented impaired recall and recognition of a critical (threat-
ening) target individual for officers experiencing reduced process-
ing capacity (as a function of exertion). However, detection and
recall of additional risk factors in the environment was not im-
paired suggesting that attentional resources may have been di-
verted to risk assessment activities rather than the encoding of an
unarmed albeit verbally threatening target (see Hope et al., 2013).
A similar explanation may apply here and is consistent with the
predictions of the arousal-based competition model (ABC; Mather
& Sutherland, 2011). Future research should examine cognitive
processing in the aftermath of weapon discharge to elucidate the
relative roles of stress, cognitive load, and competing priorities in
processing an incident and the subsequent impact on memory.
Furthermore, researchers (and, by extension, the legal profession
and evaluators) may need to a take a more contextual perspective
before assuming what constitutes “critical” stimuli in a scene. As
noted by Mather and Sutherland (2011) “priority is determined by
bottom-up perceptual salience and top-down relevance” (p. 19),
both of which would have been determined by individuals juggling
competing operational demands in the current study.
The finding that 18% of the sample (largely equivalent across
both Active and Observer witnesses) reported that the perpetrator
pointed a weapon at them in the final critical scene is interesting—
albeit predicted by expectation-driven processing. It is also con-
sistent Morgan et al.’s finding that soldiers made nontrivial errors
for the presence of weapons under stressful conditions (e.g., 27%
of soldiers falsely reported that their interrogator wielded a weap-
on; Morgan et al., 2013). However, there are at least two other
potential explanations that might account for this error. The first
concerns a relatively simple visual effect comprising two stag-
es—a feed-forward sweep that allows the rapid extraction of
features from a visual scene followed by recurrent processing to
produce a conscious experience (see Lamme, 2006). It may be that
the expectation of a weapon in combination with visual processing
under time pressure failed to detect that a weapon was not present,
leaving (some) witnesses with the belief that they had in fact seen
a weapon without the opportunity to fully process the scene.
Alternatively, it may be that the reporting of a weapon simply
reflects a reconstructive error in light of outcome knowledge. All
officers who reported seeing a weapon pointed at them also dis-
charged their own weapon. Knowing this outcome of their own
decision-making, some almost certainly felt under pressure to
justify the use of lethal force in a legally consistent manner. This
is an interesting distinction that is important for investigators to
consider in the evaluation of accounts of firearms incidents. Gen-
erally, officers will have made a decision to discharge their
weapon as a function of their contemporaneous perception and
sense-making at the scene (i.e., “What I think I see/What I think is
happening”) whereas their account justifying their actions at the
scene is likely to be based on retrospective and necessarily recon-
structive processes (i.e., “What I think I must have seen/What must
have happened”). Just as legal decisions in the aftermath of a
shooting are made with the benefit of hindsight and biases asso-
ciated with the presence of outcome information (Villejoubert,
O’Keeffe, Alison, & Cole, 2006), officers’ post hoc evaluations
may be similarly vulnerable to the influence of hindsight and
outcome information. In summary, it is not possible to determine
whether this error reflects a memory distortion or a post hoc
justification informed by outcome bias. Similarly, it cannot be
easily attributed to attentional phenomena (e.g., “inattentional
blindness”; Chabris, Weinburger, Fontaine, & Simons, 2011) in
the absence of relevant measurement data—particularly given ex-
posure duration, the error reporting rate, and distribution of the
error across active and observer conditions. Nonetheless, during
debriefing, officers who had reported seeing a weapon expressed
surprise when told that was not possible. Therefore, further re-
search is necessary to disentangle the cognitive and social effects
producing this erroneous reporting under stressful conditions. In-
deed, the current research only considered the effects of stress at
encoding on output. Given the high stakes of real-life shooting
investigations, it is reasonable to assume that officers may also
experience stress at the reporting stage.
Although officers took part in a challenging, tightly controlled,
simulated incident that replicated high-quality training scenarios,
the experience was obviously not as dangerous or consequential as
a real-world incident involving lethal weapons. Furthermore, we
only ran participants in one stimulus event. However, little re-
search in this particular applied context, with the exception of
studies conducted by Morgan and his colleagues (2013, 2014) in a
military setting, has achieved similar levels of ecological validity
(see Hope et al., 2012, 2013 for further discussion of this issue).
Our decision to recruit pairs of officers was deliberate to limit any
extraneous effects of law enforcement training and experience on
recall performance. Of course, it is possible that the performance
of the officer designated observer status does not necessarily
replicate the performance of a lay, bystander witness. Future
research should consider the extent to which expertise or domain
knowledge held by operational witnesses contributes to their sub-
sequent recall of incident and explore whether this knowledge can
be capitalized on to support retrieval for operational incidents. It
should also be noted that the current research involved a single
White male target—perpetrator race was not manipulated and nor
was “decision to shoot” a key dependent variable. Therefore, the
current research cannot speak to racial aspects of recent high
profile police shootings in the United States, or indeed elsewhere
(for recent research on race and shooting behavior, see Cox,
Devine, Plant, & Schwartz, 2014; Sim, Correll, & Sadler, 2013).
Further research, involving high fidelity simulations and method-
ologically rigorous experimental designs, is needed to examine the
determinants of decisions to shoot in diverse policing contexts.
Finally, as is common in research investigating eyewitness recall,
our analytical approach necessitated a number of statistical tests,
which can increase the likelihood of Type I (familywise) error. For
this reason, our interpretation of our results relies heavily on
measures of effect size, rather than solely the statistical signifi-
cance of any finding.
Through examining the performance of witnesses who, by virtue
of their duty, are required not only to witness but also to react and
respond under stressful conditions, the current research constitutes
an important and timely contribution both to the psychological
literature and wider policy concerns in legal and investigative
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
10 HOPE ET AL.
contexts. To date, little research has systematically examined the
recall of officers for challenging or threatening operational inci-
dents, particular those involving use of lethal force. However, the
investigation of such incidents constitutes a major and high profile
task both for police forces and external agencies such as the
Independent Police Complaints Commissions (IPCC) established
in the U.K. under the Police Reforms Act, 2002. Internationally,
such investigations are typically high profile, attracting both public
and media attention, and have serious consequences for the offi-
cers involved (Goodwill et al., 2010). Thus, the development of
evidence-based policy and investigative practice is critical. The
current results document the vulnerability of memory in this con-
text and highlight the need for well-informed approaches to elic-
iting information from operationally active witnesses.
References
ACPO (Association of Chief Police Officers; 2011). Manual of guidance
on the management, command and deployment of armed officers (3rd
ed.). Retrieved from http://www.acpo.police.uk/documents/uniformed/
2011/201111MCDofAO3.pdf
Adolphs, R., Tranel, D., & Buchanan, T. W. (2005). Amygdala damage
impairs emotional memory for gist but not details of complex stimuli.
Nature Neuroscience, 8, 512–518.
Akinola, M., & Mendes, W. B. (2012). Stress-induced cortisol facilitates
threat-related decision making among police officers. Behavioral Neu-
roscience, 126, 167–174. http://dx.doi.org/10.1037/a0026657
Alferes, V. R., & Kenny, D. A. (2009). SPSS programs for the measure-
ment of nonindependence in standard dyadic designs. Behavior Re-
search Methods, 41, 47–54. http://dx.doi.org/10.3758/BRM.41.1.47
Alpert, G. P. (2009). Interpreting police use of force and the construction
of reality. Criminology & Public Policy, 8, 111–115. http://dx.doi.org/
10.1111/j.1745-9133.2009.00536.x
Armed Policing Authorised Professional Practice. (2013). Armed policing.
Retrieved from http://www.app.college.police.uk/app-content/armed-
policing/?s⫽
Banks, J. B., Tartar, J. L., & Welhaf, M. S. (2014). Where’s the impair-
ment: An examination of factors that impact sustained attention follow-
ing a stressor. Cognition and Emotion, 28, 856–866. http://dx.doi.org/
10.1080/02699931.2013.857643
Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable
distinction in social psychological research: Conceptual, strategic, and
statistical considerations. Journal of Personality and Social Psychology,
51, 1173–1182. http://dx.doi.org/10.1037/0022-3514.51.6.1173
Benton, T. R., Ross, D. F., Bradshaw, E., Thomas, W. N., & Bradshaw,
G. S. (2006). Eyewitness memory is still not common sense: Comparing
jurors, Judges and Law Enforcement to eyewitness experts. Applied
Cognitive Psychology, 20, 115–129. http://dx.doi.org/10.1002/acp.1171
Betts, K. R., & Hinsz, V. B. (2013). Strong shared representations
promote schema-consistent memory errors in groups. Group Pro-
cesses & Intergroup Relations, 16, 734–751. http://dx.doi.org/10
.1177/1368430213486206
Braidwood, T. R. (2010). Why? The Robert Dziekanski Tragedy: Braid-
wood commmission on the death of Robert Dziekanski (B. C.). ISBN
978-0-7726-6252-1. Retrieved from http://www.braidwoodinquiry.ca/
report/P2Report.php
Chabris, C. F., Weinberger, A., Fontaine, M., & Simons, D. J. (2011). You
do not talk about Fight Club if you do not notice Fight Club: Inatten-
tional blindness for a simulated real-world assault. Iperception, 2, 150–
153. http://dx.doi.org/10.1068/i0436
Clark, R. C. (2008). Building expertise: Cognitive methods for training and
performance improvement. San Francisco, CA: Wiley.
Conway, M. A. (2012). Ten things the law and others should know about
human memory. In L. Nadel & W. Sinnott-Armstrong (Eds.), Memory
and law (pp. 359–372). New York, NY: Oxford University Press.
http://dx.doi.org/10.1093/acprof:oso/9780199920754.003.0014
Cox, W. T. L., Devine, P. G., Plant, E. A., & Schwartz, L. L. (2014).
Towards a comprehensive understanding of officers’ shooting decisions:
No simple answers to this complex problem. Basic and Applied Social
Psychology, 36, 356–364. http://dx.doi.org/10.1080/01973533.2014
.923312
Davis, D., & Loftus, E. F. (2009). Expectancies, emotion and memory
reports of visual events: Applications in the legal system. In J. R.
Brockmole (Ed.), The visual world in memory (pp. 178–214). New
York, NY: Psychology Press.
Deffenbacher, K. A., Bornstein, B. H., Penrod, S. D., & McGorty, E. K.
(2004). A meta-analytic review of the effects of high stress on eyewit-
ness memory. Law and Human Behavior, 28, 687–706. http://dx.doi.org/
10.1007/s10979-004-0565-x
De Kloet, E. R., Vreugdenhil, E., Oitzl, M. S., & Joëls, M. (1998). Brain
corticosteroid receptor balance in health and disease. Endocrine Re-
views, 19, 269–301.
deq Uervain, D. J., Roozendaal, B., Nitsch, R. M., McGaugh, J. L., &
Hock, C. (2000). Acute cortisone administration impairs retrieval of
long-term declarative memory in humans. Nature Neuroscience, 3, 313–
314. http://dx.doi.org/10.1038/73873
Dodd, V., & Travis, A. (2014). Mark Duggan police must stop refusing
to be interviewed, says IPCC head. Retrieved from http://www
.theguardian.com/uk-news/2014/jan/14/mark-duggan-police-must-talk-
says-ipcc
Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the
organization of behavior. Psychological Review, 66, 183–201. http://dx
.doi.org/10.1037/h0047707
Eberhardt, J. L., Goff, P. A., Purdie, V. J., & Davies, P. G. (2004). Seeing
black: Race, crime, and visual processing. Journal of Personality and
Social Psychology, 87, 876–893. http://dx.doi.org/10.1037/0022-3514
.87.6.876
Engelhard, I. M., van den Hout, M. A., & McNally, R. J. (2008). Memory
consistency for traumatic events in Dutch soldiers deployed to Iraq.
Memory, 16, 3–9. http://dx.doi.org/10.1080/09658210701334022
Eyre, M., & Alison, L. (2007). To decide or not to decide: Decision making
and decision avoidance in critical incidents. In D. Carson, R. Milne, F. J.
Pakes, K. Shalev, & A. Shawyer (Eds.), Applying psychology to criminal
justice (pp. 211–232). Chichester: Wiley. http://dx.doi.org/10.1002/
9780470713068.ch13
Fisher, R. P., & Geiselman, R. E. (1992). Memory enhancing techniques
for investigative interviewing: The cognitive interview. Springfield, IL:
Charles C Thomas.
Goodwill, A. M., Alison, L., Lehmann, R., Francis, A., & Eyre, M. (2010).
The impact of outcome knowledge, role, and quality of information on
the perceived legitimacy of lethal force decisions in counter-terrorism
operations. Behavioral Sciences & the Law, 28, 337–350.
Healy, A. F., Kole, J. A., & Bourne, L. E., Jr. (2014). Training principles
to advance expertise. Frontiers in Psychology, 5, 131. http://dx.doi.org/
10.3389/fpsyg.2014.00131
Heaton-Armstrong, A., & Wolchover, D. (1993, April). Knocking heads
together. Counsel, 14–15.
Heaton-Armstrong, A., & Wolchover, D. (2009, January). Conferring at
the crossroads. Counsel, 18.
Hockey, G. R. J. (1993). Cognitive-energetical control mechanisms in the
management of work demands and psychological health. In A. D.
Baddeley & L. Weiskrantz (Eds.), Attention, selection. Awareness and
control: A tribute to Donald Broadbent (pp. 328–345). Oxford: Oxford
University Press.
Hockey, G. R. J. (1997). Compensatory control in the regulation of human
performance under stress and high workload; a cognitive-energetical
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
11
MEMORY AND THE OPERATIONAL WITNESS
framework. Biological Psychology, 45, 73–93. http://dx.doi.org/10
.1016/S0301-0511(96)05223-4
Hockey, G. R. J., & Hamilton, P. (1983). The cognitive patterning of stress.
In G. R. J. Hockey (Ed.), Stress and fatigue in human performance (pp.
141–177). New York, NY: Wiley.
Hope, L., Gabbert, F., & Fraser, J. (2013). Postincident conferring by law
enforcement officers: Determining the impact of team discussions on
statement content, accuracy, and officer beliefs. Law and Human Be-
havior, 37, 117–127. http://dx.doi.org/10.1037/lhb0000019
Hope, L., Lewinski, W., Dixon, J., Blocksidge, D., & Gabbert, F. (2012).
Witnesses in action: The effect of physical exertion on recall and
recognition. Psychological Science, 23, 386–390. http://dx.doi.org/10
.1177/0956797611431463
Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic data analysis.
New York, NY: Guilford Press.
Kleider, H. M., Pezdek, K., Goldinger, S. D., & Kirk, K. A. (2008).
Schema-driven source misattribution errors: Remembering the expected
from a witnessed event. Applied Cognitive Psychology, 22, 1–20. http://
dx.doi.org/10.1002/acp.1361
Klein, G. A. (1998). Sources of power: How people make decisions.
Cambridge, MA: MIT Press.
Kruglanski, A. W., & Freund, T. (1983). The freezing and un-freezing of
layinferences: Effects of impressional primacy, ethnic stereotyping and
numerical anchoring. Journal of Experimental Social Psychology, 19,
448–468. http://dx.doi.org/10.1016/0022-1031(83)90022-7
Kuhlmann, S., Piel, M., & Wolf, O. T. (2005). Impaired memory retrieval
after psychosocial stress in healthy young men. The Journal of Neuro-
science, 25, 2977–2982. http://dx.doi.org/10.1523/JNEUROSCI.5139-
04.2005
Lamb, M. E., Orbach, Y., Hershkowitz, I., Esplin, P. W., & Horowitz, D.
(2007). A structured forensic interview protocol improves the quality
and informativeness of investigative interviews with children: A review
of research using the NICHD Investigative Interview Protocol. Child
Abuse & Neglect, 31, 1201–1231. http://dx.doi.org/10.1016/j.chiabu
.2007.03.021
Lamme, V. A. F. (2006). Towards a true neural stance on consciousness.
Trends in Cognitive Sciences, 10, 494–501. http://dx.doi.org/10.1016/j
.tics.2006.09.001
Landis, J. R., & Koch, G. G. (1977). The measurement of observer
agreement for categorical data. Biometrics, 33, 159–174. http://dx.doi
.org/10.2307/2529310
Leach, J., & Ansell, L. (2008). Impairment in attentional processing in a
field survival environment. Applied Cognitive Psychology, 22, 643–652.
http://dx.doi.org/10.1002/acp.1385
Leach, J., & Griffith, R. (2008). Restrictions in working memory capacity
during parachuting: A possible cause of ‘no pull’ fatalities. Applied
Cognitive Psychology, 22, 147–157. http://dx.doi.org/10.1002/acp.1364
Lommen, M. J. J., Engelhard, I. M., & van den Hout, M. A. (2013).
Susceptibility to long-term misinformation effect outside of the labora-
tory. European Journal of Psychotraumatology, 4, 19864. http://dx.doi
.org/10.3402/ejpt.v4i0.19864
Lupien, S. J., & Lepage, M. (2001). Stress, memory, and the hippocampus:
Can’t live with it, can’t live without it. Behavioural Brain Research,
127, 137–158. http://dx.doi.org/10.1016/S0166-4328(01)00361-8
Mather, M., & Sutherland, M. R. (2011). Arousal-biased competition in
perception and memory. Perspectives on Psychological Science, 6, 114–
133. http://dx.doi.org/10.1177/1745691611400234
McGaugh, J. L. (2013). Making lasting memories: Remembering the
significant. Proceedings of the National Academy of Sciences of the
United States of America, 110, 10401–10407. http://dx.doi.org/10.1073/
%2Fpnas.1301209110
Meyerhoff, J. L., Norris, W., Saviolakis, G. A., Wollert, T., Burge, B.,
Atkins, V., & Spielberger, C. (2004). Evaluating performance of law
enforcement personnel during a stressful training scenario. Annals of the
New York Academy of Sciences, 1032, 250–253. http://dx.doi.org/10
.1196/annals.1314.031
Morgan, C. A., III, Hazlett, G., Doran, A., Garrett, S., Hoyt, G., Thomas,
P., . . . Southwick, S. M. (2004). Accuracy of eyewitness memory for
persons encountered during exposure to highly intense stress. Interna-
tional Journal of Law and Psychiatry, 27, 265–279. http://dx.doi.org/10
.1016/j.ijlp.2004.03.004
Morgan, C. A., III, & Southwick, S. (2014). Perspective: I believe what I
remember, but it may not be true. Neurobiology of Learning and Mem-
ory, 112, 101–103. http://dx.doi.org/10.1016/j.nlm.2013.12.011
Morgan, C. A., III, Southwick, S., Steffian, G., Hazlett, G. A., & Loftus,
E. F. (2013). Misinformation can influence memory for recently expe-
rienced, highly stressful events. International Journal of Law and Psy-
chiatry, 36, 11–17. http://dx.doi.org/10.1016/j.ijlp.2012.11.002
Morley, J. E., & Farr, S. A. (2012). Hormesis and amyloid-protein:
Physiology or pathology? Journal of Alzheimer’s Disease, 29, 487–492.
Most, S. B., Scholl, B. J., Clifford, E. R., & Simons, D. J. (2005). What you
see is what you set: Sustained inattentional blindness and the capture of
awareness. Psychological Review, 112, 217–242. http://dx.doi.org/10
.1037/0033-295X.112.1.217
Neisser, U. (1976). Cognition and reality. San Francisco, CA: Freeman.
Nieuwenhuys, A., Savelsbergh, G. J. P., & Oudejans, R. R. (2012). Shoot
or don’t shoot? Why police officers are more inclined to shoot when they
are anxious. Emotion, 12, 827–833. http://dx.doi.org/10.1037/a0025699
Payne, J. D., Jackson, E. D., Ryan, L., Hoscheidt, S., Jacobs, J. W., &
Nadel, L. (2006). The impact of stress on neutral and emotional aspects
of episodic memory. Memory, 14, 1–16. http://dx.doi.org/10.1080/
09658210500139176
Payne, J. D., Nadel, L., Allen, J. J., Thomas, K. G., & Jacobs, W. J. (2002).
The effects of experimentally induced stress on false recognition. Mem-
ory, 10, 1–6. http://dx.doi.org/10.1080/09658210143000119
Phelps, E. A. (2006). Emotion and cognition: Insights from studies of the
human amygdala. Annual Review of Psychology, 57, 27–53. http://dx
.doi.org/10.1146/annurev.psych.56.091103.070234
Pickel, K. L. (1999). The influence of context on “weapon focus” effect.
Law and Human Behavior, 23, 299–311. http://dx.doi.org/10.1023/A:
1022356431375
Plant, K., & Stanton, N. A. (2012). Why did the pilots shut down the wrong
engine? Explaining errors in context using Schema Theory and the
Perceptual Cycle Model. Safety Science, 50, 300–315. http://dx.doi.org/
10.1016/j.ssci.2011.09.005
Qin, S., Hermans, E. J., van Marle, H. J. F., & Fernández, G. (2012).
Understanding low reliability of memories for neutral information en-
coded under stress: Alterations in memory-related activation in the
hippocampus and midbrain. The Journal of Neuroscience, 32, 4032–
4041. http://dx.doi.org/10.1523/JNEUROSCI.3101-11.2012
Roozendaal, B. (2000). Glucocorticoids and the regulation of memory
consolidation. Psychoneuroendocrinology, 25, 213–238. http://dx.doi
.org/10.1016/S0306-4530(99)00058-X
Safer, M. A., Christianson, S.-Å., Autry, M. W., & Österlund, K. (1998).
Tunnel memory for traumatic events. Applied Cognitive Psychology, 12,
99–117. http://dx.doi.org/10.1002/(SICI)1099-0720(199804)12:2⬍99::
AID-ACP509⬎3.0.CO;2-7
Schacter, D. L. (1999). The seven sins of memory. Insights from psychol-
ogy and cognitive neuroscience. American Psychologist, 54, 182–203.
http://dx.doi.org/10.1037/0003-066X.54.3.182
Schoofs, D., Wolf, O. T., & Smeets, T. (2009). Cold pressor stress impairs
performance on working memory tasks requiring executive functions in
healthy young men. Behavioral Neuroscience, 123, 1066–1075. http://
dx.doi.org/10.1037/a0016980
Scoboria, A., & Fisico, S. (2013). Encouraging and clarifying “don’t
know” responses enhances interview quality. Journal of Experimental
Psychology: Applied, 19, 72–82. http://dx.doi.org/10.1037/a0032067
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
12 HOPE ET AL.
Shackman, A. J., Sarinopoulos, I., Maxwell, J. S., Pizzagalli, D. A., Lavric,
A., & Davidson, R. J. (2006). Anxiety selectively disrupts visuospatial
working memory. Emotion, 6, 40–61. http://dx.doi.org/10.1037/1528-
3542.6.1.40
Sharman, S. J., & Powell, M. B. (2012). Comparison of adult witnesses’
suggestibility across various types of leading questions. Applied Cogni-
tive Psychology, 26, 48–53. http://dx.doi.org/10.1002/acp.1793
Sherman, J. W., & Bessenoff, G. R. (1999). Stereotypes as source moni-
toring cues: On the interaction between episodic and semantic memory.
Psychological Science, 10, 106–110. http://dx.doi.org/10.1111/1467-
9280.00116
Sim, J. J., Correll, J., & Sadler, M. S. (2013). Understanding police and
expert performance: When training attenuates (vs. exacerbates) stereo-
typic bias in the decision to shoot. Personality and Social Psychology
Bulletin, 39, 291–304. http://dx.doi.org/10.1177/0146167212473157
Simons, D. J., & Chabris, C. F. (2011). What people believe about how
memory works: A representative survey of the U.S. population. PLoS
ONE, 6, e22757. http://dx.doi.org/10.1371/journal.pone.0022757
Spierer, D. K., Griffiths, E., & Sterland, T. (2009). Fight or flight: Mea-
suring and understanding human stress response in tactical situations.
The Tactical Edge. Retrieved from http://www.polar.com/files/pdf/
Flight_or_Fight_-_Using_Heart_Rate_and_Heart_Rate_Variability_to_
Measure_Bomb_Technician_Stress.pdf
Taverniers, J., Taylor, M. K., & Smeets, T. (2013). Delayed memory
effects after intense stress in Special Forces candidates: Exploring path
processes between cortisol secretion and memory recall. Stress, 16,
311–320. http://dx.doi.org/10.3109/10253890.2012.721824
Thayer, J. F., Ahs, F., Fredrikson, M., Sollers, J. J., III, & Wager, T. D.
(2012). A meta-analysis of heart rate variability and neuroimaging
studies: Implications for heart rate variability as a marker of stress and
health. Neuroscience and Biobehavioral Reviews, 36, 747–756. http://
dx.doi.org/10.1016/j.neubiorev.2011.11.009
Tuckey, M. R., & Brewer, N. (2003). The influence of schemas, stimulus
ambiguity, and interview schedule on eyewitness memory over time.
Journal of Experimental Psychology: Applied, 9, 101–118. http://dx.doi
.org/10.1037/1076-898X.9.2.101
Villejoubert, G., O’Keeffe, C. J., Alison, L. J., & Cole, J. C. (2006).
Hindsight bias and shooting incidents. In S. Giles & M. Santarcangelo
(Eds.), Psychological aspects of legal processes (pp. 17–24). Liverpool:
IA-IP Press.
Vrij, A., Hope, L., & Fisher, R. P. (2014). Eliciting reliable information in
investigative interviews. Policy Insights from Behavioral and Brain
Sciences, 1, 129–136. http://dx.doi.org/10.1177/2372732214548592
Wolf, O. T. (2003). HPA axis and memory. Best Practice & Research
Clinical Endocrinology & Metabolism: Clinical and Experimental, 17,
287–299. http://dx.doi.org/10.1016/S1521-690X(02)00101-X
Wolf, O. T., Schommer, N. C., Hellhammer, D. H., McEwen, B. S., &
Kirschbaum, C. (2001). The relationship between stress induced cortisol
levels and memory differs between men and women. Psychoneuroen-
docrinology, 26, 711–720. http://dx.doi.org/10.1016/S0306-
4530(01)00025-7
Woody, E. Z., & Szechtman, H. (2011). Adaptation to potential threat: The
evolution, neurobiology, and psychopathology of the security motivation
system. Neuroscience and Biobehavioral Reviews, 35, 1019–1033.
http://dx.doi.org/10.1016/j.neubiorev.2010.08.003
Received September 17, 2014
Revision received August 14, 2015
Accepted August 16, 2015 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
13
MEMORY AND THE OPERATIONAL WITNESS
A preview of this full-text is provided by American Psychological Association.
Content available from Law and Human Behavior
This content is subject to copyright. Terms and conditions apply.